Apple 提出 SELFCOMPACT：让模型自己决定何时压缩长智能体轨迹
Self-Compacting Language Model Agents
Apple 给出了一种新的长任务智能体压缩方案：不再按固定 token 数截断，而是让模型自己调用工具决定何时总结，避免丢掉中间结果。
Apple 研究团队提出 SELFCOMPACT 脚手架，针对长智能体轨迹中推理链与工具调用累积的陈旧内容问题。现有方法在 token 阈值处定时压缩，可能丢弃推导或搜索中途的部分结果。SELFCOMPACT 将触发时机交给模型，配合一个压缩工具让模型主动总结累积轨迹。两个元素均为推理时组件，不改变底层模型。
Self-Compacting Language Model Agents
Long agent traces composed of chains of thought and tool calls accumulate stale content that anchor subsequent generations, and eventually outgrow the context window. Existing scaffolds mitigate it with fixed-interval compaction triggered at a token threshold. Such triggers pay no heed to trajectory structure, risking discard of partial results mid-derivation or mid-search. We propose SELFCOMPACT, a scaffold that allows the model itself to decide when and how to compact. Specifically, it pairs two inference-time elements: (i) a compaction tool the model invokes to summarize the accumulated…
